Activation rate is the share of new installs that complete one chosen first-value action within a fixed window. Divide activated installs by installs: 1,000 installs and 280 activated within 7 days is 28%. The hard part is choosing the action and the window.
Activation rate formula
Activation rate = installs that fired the activation event within the window ÷ installs in the cohort.
Three parts need a fixed definition before you measure anything:
- Cohort: the installs from one period, counted once per device at first open.
- Activation event: the single action that means the user got the core value.
- Window: how long after install the event still counts, for example 7 days.
The numerator comes only from devices in the cohort. A user who installed last month and fires the event today belongs to last month's cohort, not this one.
Measure a cohort only after its window has closed. With a 7-day window, yesterday's installs have not had the chance to activate, so their rate is incomplete and will rise.
How to track activation rate in ActivationPal
ActivationPal records one first_open event per install, and every funnel stage counts unique devices from that install cohort. No stage can exceed 100% of installs. That gives you the denominator without any setup.
Which path you take depends on what you chose as the activation event.
Activation is finishing onboarding. Call onboardingCompleted() at the end of the flow. The Onboarded headline and the funnel's completed stage show it against installs.
Activation is an action inside the product. Send it as a custom event, for example ActivationPal.track("workout_logged", ["type": "run"]). The event needs no registration: it appears in the Events card with its count and its unique devices.
The Events card counts devices that sent the event in the selected date range, including people who installed earlier. For a clean cohort rate, divide only by devices that installed in the same period. Or pull raw rows with get_events (each has name, ts and device_id) and compare every device's event time with its first_open. The full event list is in the events docs; the product overview is on the homepage.
What counts as an activation event
An activation metric is only as good as the event behind it. The event is a user action that delivers the core value, not a screen view and not a tap on a button that leads nowhere.
By app type:
- Fitness or habit app: first logged workout or completed session. The user has put real effort into the product.
- Content or media app: first item played past a minimum length. Opening a title is browsing, finishing a part of it is use.
- Utility or scanner app: first successful result saved, such as the first scan. The tool did its job once.
- Social or dating app: first message sent or first match made. Value needs another person.
- Finance or tracker app: first entry logged or first account connected. Without data the app shows nothing.
Name it in snake_case and past tense, like scan_saved. Do not use the paywall or the purchase: that is conversion, covered in app conversion rate.
To test a candidate, compare the retention of devices that fired it with devices that did not, using the method in mobile app retention rate. If the event does not separate the two groups, it is not your activation event.
Activation window
Pick the window by how soon first value is possible. A utility that returns a result in one session can use day 0. A habit app where value builds over a few days fits 7 days.
Start the window at first open. Fix it before you look at results, use the same one for every cohort, and count a device once, at its first activation event. Changing the window between cohorts makes the trend meaningless.
Activation rate vs onboarding completion
Onboarding completion measures whether a user finished the setup flow. Activation measures whether they got value.
A user can finish every onboarding screen and never log a workout. Another can skip onboarding and activate in the first minute. Treating the two as one hides both cases.
Onboarding completion works as the activation event only when the onboarding itself delivers the value, such as a quiz that ends in a personal plan. Otherwise track both and read them side by side. The onboarding completion rate article covers how to count the first one.
Find where installs fail to activate
List the steps a new user passes between first_open and your activation event. Get a user count for each step, for example from the Events card or get_events. Enter the step names and counts in the funnel calculator and read the step with the largest drop-off: that is where installs stop short of value.